The goal of this research is to dissect the causative genes and mutations underlying a rapidly evolving complex trait and to understand the nature of constraints on the evolution of this trait. Such an endeavor requires an experimental system in which drastic changes occur over short evolutionary timeframes in species that can be easily manipulated genetically. The posterior lobe in the Drosophila melanogaster species group is a recently-derived structural feature of male genitalia important in mating. The structure has evolved strikingly different morphology in each of the four species of this group over the past two million years. Our preliminary work has documented the apparent co-option and re-deployment of a highly conserved gene regulatory network (GRN) encoding the larval posterior spiracle in creating novel phenotypic variation in the adult posterior lobe. This finding stimulates us to hypothesize mechanisms that would allow the rapidly evolving lobe to change while leaving the spiracle structure unaltered. Among the members of this GRN is the gene pox neuro (poxn) which we have shown is divergently upregulated in D. simulans relative to D. melanogaster and that over-expression in D. melanogaster leads to enlargement of the lobe as observed in D. simulans. Additionally, we have shown that the upregulation of poxn in D. simulans is caused by sequence divergence in a posterior lobe-specific enhancer contained within the gene. Using a population genetic method to detect natural selection, we have identified a signature of selection that coincides with the posterior lobe enhancer of poxn. Here, we propose to employ genomic (i.e. RNA-seq, population genomics) and molecular genetic techniques (i.e. transgenic expression and complementation assays) to identify other causative genes and mutations responsible for differences in posterior lobe size and shape among species of the D. melanogaster species group (Aims 1 and 2). With a large collection of such molecular changes in hand, we will examine how co-opted nodes of the posterior lobe have diverged in their lobe functions without altering the spiracle structure (Aim 3).